On Another Theory of Marxian Exploitation with Oscar Lange and Paul Samuelson

In a forthcoming review by my friend, M.A. ‘Ali’ Khan (Hopkins), I marked some quoted passages from Paul. A. Samuelson (1967) "The monopolistic competition revolution" Monopolistic competition theory: Studies in impact, essays in honor of Edward H. Chamberlin. My references are to the page-numbers of this Festschrift. Thanks to Ali, I was able to get my hands on a pdf copy. (It’s been reprinted (inter alia) here.) Samuelson’s (1967) paper is a veritable tour de force, and I lack the expertise to summarize it adequately.

The concern with exploitation is introduced obliquely in a footnote (11) that I quote in order to illustrate Samuelson’s tendency to juggle multiple conceptual and intellectual balls at once:

Even within the constant returns technology, we shall see that there are possible advantages (and no disadvantages!) to be derived from having the owners of any unique factor of production, e.g., land suitable for mulberry growing, form a coalition that exploits its monopoly power. We should not wonder that the calculating self-interest on which Adam Smith relies to move the Invisible Hand of perfect competition should motivate people to utilize the ballot box of democracy to institute crop control and other public programs interfering with perfect competition. To a psychologist, Bentham’s individualism and Webb’s Fabianism are one in motive and appeal. (p. 115—ignore the way he presents Adam Smith. He knows better.)

Here exploitation is an effect of monopoly power, especially of a (relative fixed) factor of production. But monopoly also presupposes politics — or effective coordination — because it involves the ‘formation’ (and maintenance) of a “coalition.” And the effect of exploitation, in turn, is advantage to the class participating in the coalition, that is, “the owners” of that factor. So, here a regime of property rights and enforcement is also presupposed.

As an aside, my renewed interest was prompted by a fruitful workshop on Monday in Nottingham on “The Idea of Freedom and Modern Slavery” organized by Diane Popescu (Nottingham). There Gulshan Khan (Nottingham) defended the idea (to simplify) that modern slavery just is intense and extensive exploitation. This resonated with some remarks I had made (while commenting on Du Bois in my presentation) on the fact that since wage-labor is often treated as a kind of slavery within Marxism, it’s clear that for this tradition the difference between ordinary work and slavery is a matter of the intensity of exploitation.

Be that as it may, a bit later, Samuelson introduces a digression: “Let me digress to recognize that there are other causes of monopolistic imperfection than deviations from constant returns.” (p. 117) The third of such causes is introduced with some old-fashioned terminology, and I will quote the whole paragraph (and then the note attached to it).

The old-fashioned notion of contrived scarcity, as distinguished from natural scarcity, deserves mention. Reformers used to speak of the “land monopoly,” meaning no more than that land is scarce, and earns very high incomes for the few people who happen to own much of it. Fabians like Shaw and Webb also used such terms, and extended this concept of monopoly to all private property (at the same time making what we would call the Austrian assumption that all factors of production can be regarded as being fairly inelastic in supply to society and hence can be viewed as “rent-earning”). Often they joined with Marxians in regarding human labor as being exploited in the sense of getting less than 100 percent of the national product, having to give up too large a fraction to undeserving owners of property.

So, it turns out that the passage from p. 115 discussed above has an old-fashioned and (ahh) up-to-date kind of analysis. In fact, we can discern in the old-fashioned notion Adam Smith’s Wealth of Nations (see 1.7.24-26, p. 78 in Glasgow edition). The socialist Fabians, who like to think of themselves as up-to-date, actually are more Smithian than they wish to realize.

Joking aside, here a distinct notion of exploitation is introduced. This involves (to simplify) the reduction on the income of workers caused by the (mere) existence of property. I ignore that ‘too large a fraction’, but see below.

The accompanying footnote (reads as follows:

In the 1930’s, Oskar Lange urged Marxists to look at the problem of exploitation in this way and not to waste time on the orthodox Marxian labor theory of value. 0. Lange, “Marxian Economics and Modern Economic Theory,” Review of Economic Studies, II (1935), pp. 189-201. For essentially the same view, see J. Robinson, An Essay on Marxian Economics, Macmillan, London, 1942. For a diagnosis of property income that neglects the identification made by such reformers of nonland property with land property, see G. J. Stigler, “Bernard Shaw Sidney Webb, and the Theory of Fabian Socialism,” Proceedings of the American Philosophical Society, CIII (1959), pp. 469-475. (p. 118 n. 14)

Now, Lange’s (1930 paper (here) is itself a tour de force. One that I warmly recommend to my young readers because it does a very fine job at offering a comparative assessment of the strength and weaknesses of Marxian economics relative to non-Marxian schools of thought. Samuelson has read it carefully because the passage he has in mind is itself an extended footnote in it:

In the Marxian system the labour theory of values serves also to demonstrate the exploitation of the working class under Capitalism, i.e. the difference between the personal distribution of income in a capitalist economy and in an "einfache Warenproduktion" It is this deduction from the labour theory of value which makes the orthodox Marxist stick to it. But the same fact of exploitation can also be deduced without the help of the labour theory of value. Also without it, it is obvious that the personal distribution of income in a capitalist economy is different from that in an "einfache Warenproduktion" (or in a socialist economy based on equalitarian principles, in which the distribution of income would be substantially the same as in an "einfache Warenproduktion"), for profit, interest and rent can obviously be the personal income of a separate class of people only in a capitalist economy. If interest is explained by the marginal productivity of capital, it is only because the workers do not own the capital they work with that interest is the personal income of a separate class of people. If interest is regarded as due to a higher valuation of present than future goods it is only because the workers do not possess the subsistence fund enabling them to wait until the commodities they produce are ready that the capitalist advancing it to the workers gets the interest as his personal income. Just as in Marx's case it is because the workers do not possess the means of production that the surplus value is pocketed by the capitalist. To make the Marxian concept of exploitation clearer by contrast it may be noticed that Pigou (The Economics of Welfare, 3rd ed., 1929, p. 556) and Mrs. Robinson (The Economics of Imperfect Competition, p. 28I seq.) define exploitation of the worker as occurring when he gets less than the value of the marginal physical product of his labour. This means that exploitation is defined by contrasting the distribution of income in monopolistic Capitalism and in competitive Capitalism. The middle-class character of this idea of social justice is obvious. For the Socialist the worker is exploited even if he gets the full value of the marginal product of his labour, for from the fact that interest or rent is determined by the marginal productivity of capital or land it does not follow, from the socialist point of view, that the capital- or land-owner ought to get it as his personal income. The Marxian definition of exploitation is derived from contrasting the personal distribution of income in a capitalist economy (irrespective of whether monopolistic or competitive) with that in an "einfache Warenproduktion" in which the worker owns his means of production. (p. 195, n. 3)

The Pigou/Robinson notion of exploitation is now known as the “neo classical account.” (See Zwolinski here.) I leave it aside. Lange’s own view is worth attention because it is not well remembered today (as judged by Zwolinski’s overview of Marxian accounts of exploitation.) Because Lange’s own account does not rely on the labor theory of value nor on the ‘forced appropriation of surplus value’ (which is associated with G.A. Cohen’s Marxian reinterpretation of exploitation).

It is fair to say that Samuelson subtly misrepresent Lange. Rather than urging his readers to avoid wasting “time on the orthodox Marxian labor theory of value,” Lange show how the non-Orthodox Marxist (and non-Marxist) can have a notion of exploitation that is supposed to be equivalent to the one that accepts the labor theory of value.* Lange is not rejecting the labor theory of value as such.

In ‘simple commodity production’ [einfache Warenproduktion] of the sort in which workers own the means of production, there would be no exploitation (and capitalists would derive no rents or profits). So, exploitation is measured or estimated by the sum between the income that workers would have in such an economy and the income they receive under really existing capitalism. Samuelson’s way of conveying this is not false, but obscures Lange’s own presentation.

Notice that while Lange would not deny that his account of exploitation de facto involves forced appropriation, but it is conceptually kind of independent of it (and so distinct from the Cohen approach). All he requires is that a model of simple commodity production is conceptually available and that incomes in it can be estimated.

This problem of measuring exploitation is pretty equivalent to the challenges facing Smith’s attempt to estimate natural prices in what the Marxist calls ‘classical economics.’ It’s no small problem (with that counter-factual), but not impossible to generate estimates of exploitation (which may explain Samuelson’s odd use of ‘too large a fraction.’)

In the paragraph above, I used ‘kind of’ because when we look at the small print of Lange’s approach, we see that force is not wholly neglected in his definition of capitalism:

Capitalism means an exchange-economy with private ownership of the means of production, to which the further sociological datum is added that the population is divided into two parts, one of which owns the means of production while the other part, owning no means of production, is compelled to work as wage-earners with the means of production belonging to the other part. Only because of this sociological datum do profit and interest appear as personal income separate from wages. (p. 189, n. 4; [emphasis added]

Here compulsion turns workers into kind of wage-slaves. But the sociological dictum is not intrinsic to the definition.

The material from Lange also helps explain why the “modern theorist” (Samuelson, p. 118) can discard the “old-fashioned” notions about different kinds of monopoly. For, “All the phenomena of exploitation described are completely compatible with the most pure competition.” (Samuelson, p. 118) This accurately conveys Lange’s insight against the neo-classical interpretation of exploitation (in which, as Zwolinski also notes, exploitation seems to be impossible under pure competition).+

---

*I think what’s happened here is that Samuelson has imposed his own explicit adherence to Kuhnianism onto Lange’s essay, and so has created an incommensurability where there is none. But it’s also possible that Samuelson missed that Lange is rejecting Pigou and Robinson here.

+Samuelson goes on to offer a different argument, too. But I ignore that here.